Heats of Formation and Heat Capacities in the System Iron-Nickel-Chromium 0.KUBASCHEWSKI
and L. E. H. STUART
Metallurgy Division, National Physical Laboratory, Teddington, Middlesex, England
The heats of formation from the component metals of 47 alloys in the system iron-nickel-chromium have been measured at about 1300' C. using an adiabatic high-temperature calorimeter. The present results together with the previously published heats of formation in the binary systems have been used to interpolate the isoenthalpy contours in the ternary system. A maximum heat absorption of 2200 cal. per gram-atom was found at a composition of about 10 atomic % Fe, 6 0 atomic O h Cr, and 30 atomic % Ni. The true heat capacities of a nickel-chromium alloy with 35 atomic Yo Cr and of the component metals have also been measured between 850' and 1220'C. The deviation from Neumann-Kopp's rule is positive, of the order of ACp = 0.7-0.4 cal. per O C . gram-atom. The heat of the orderdisorder transformation of an alloy, FeNia, was 2 667 cal. per gram-atom.
THIS
LABORATORY has started an investigation of the thermodynamic properties of the system iron-nickelchromium. So far, the determination of the integral heats of formation at about 1300°C. has been completed. Some heat capacities in the binary systems have also been measured, and vapor pressure and e.m.f. measurements have begun. However, since a complete study of the free energies in the ternary system will require considerable time, it is believed to be desirable to put the experimental results so far obtained on record. The heats of formation of the binary alloys have previously been determined in this laboratory by Dench ( 2 ) , who devised the adiabatic calorimeter which has been used in the present investigation. EXPERIMENTAL
The high-temperature calorimeter has been described in detail by Dench ( 2 ) . A specimen made of compacted metal powders may be held under adiabatic conditions a t elevated temperature, and measured quantities of energy may be supplied to the specimen via an internal heater. Since the publication of the first paper (21, the control system of the calorimeter has been modified by the addition of an automatic controller to replace the hand-operated furnace control resistances. The purpose of the controller is to maintain adiabatic conditions during the heating of the specimen through the reaction range (ca. 40 minutes) and also during the period of time when the reaction is going to completion at high temperature. The controller is operated from the differential thermocouples, two forms of control being applied. One controls the mean power dissipated in the furnace winding; the other superimposes small increases or decreases of power on this mean level during a given small time interval and thus has the effect of adding to or subtracting from the system small pulses of energy while maintaining the mean power a t an approximately constant level. A fuller description of this semiautomatic controller has been given by Dench ( I ) . Electrolytic chromium flake was ball-milled for 8 hours and then sieved, the fraction between 300- and 400-mesh being used. Carbonyl iron powder was reduced with dried high-purity hydrogen at 400°C. for 45 minutes. The partly sintered powder was ground and sieved, the fraction less than 400-mesh being used. Carbonyl nickel powder was prepared in the same way as the iron but received hydrogen treatment for 30 minutes a t 450" C. Vacuum fusion analyses 41 8
for oxygen and nitrogen made at this stage gave the following results: iron, 0.07% 0 2 and 0.02% N P ;nickel, 0.04% 0 2 and 0.007% N P ; chromium, 0.45% 0 2 and 0.015% N2. The compacted mixture of the metal powders was enclosed in a double-walled capsule made from 0.05-mm. tantalum sheet. At certain compositions, where the tantalum tended to alloy with the compact, the inner wall was made from 0.05-mm. tungsten sheet. I n each case, a specimen prepared as indicated above was initially heated in the calorimeter to about 400°C. using the outer furnace only. The specimen was then brought to the starting temperature of 503.5"C. using the internal heater and automatic control. This is approximately the highest temperature to which the specimen may be heated without the reaction proceeding a t an appreciable rate. The specimen was held in equilibrium at this temperature for about 10 minutes and was then heated through the reaction range to 1291.7"C. by means of the internal heater, adiabatic conditions being maintained by means of the automatic controller. At 1291.7' C. the internal heater was switched off and the heat input recorded. Adiabatic conditions were maintained until it was evident that the reaction had gone to completion. This took up to one hour and was indicated by the attainment of either a constant temperature in the specimen or of a small constant rate of change of temperature owing to small errors in the differential thermocouple system. The change in heat content of the pure component metals, the capsule, and the calorimeter was measured over the same range of temperatures in separate experiments. Their mean heat capacities in the range 1274" to 1310°C. were also measured. The heat of formation was obtained by subtracting the total change in heat content from the energy put into the specimen by the internal heater during the reaction. A binary nickel-chromium alloy (35 atomic 7% Cr) was selected on which heat capacities were measured in the range 850" to 1250°C. in 100" intervals. Heat capacities were also measured in the same temperature range on a composite specimen made up of two cylinders of pure nickel and two cyclinders of chromium, the metals being in the same atomic proportion as in the alloy. The use of the calorimeter for Cp measurements has been described by Dench and Kubaschewski ( 3 ) who investigated pure iron. Finally, the heat of transformation of an iron-nickel alloy of composition FeNis, which forms an ordered structure below 500" C., was determined in the calorimeter. Specimens JOURNAL OF CHEMICAL AND ENGINEERING DATA
Cr
containing 25.0 atomic 70 F e were alloyed a t 1000°C. for several hours and then annealed (as indicated in Table 111) to form a superlattice. The order-disorder transformation occurred over the range 540" to 58WC. At this temperature the sheathed thermocouple normally used did not give an accurate measurement of the temperature of the specimen and so a thermocouple making direct contact with the specimen was used, being tied to i t with tantalum wire. The specimens were heated through three successive temperature intervals, arranged so that the transformation occurred in the middle interval. The other intervals acted as a check that the transformation did in fact occur completely in the middle interval. The calorimeter was then cooled down and the specimen reheated through the same three intervals, the difference in the input energy required giving the heat of transformation. RESULTS AND DISCUSSION The measured integral heats of formation of the ternary alloys from the component metals are recorded in Table I. Recent results on some binary iron-nickel alloys are included. The accuracy of such measurements with the present calorimeter has been discussed by Dench ( 2 ) . The maximum error in the individual heats of alloying may be taken to be h 1 5 0 cal. per gram-atom. This includes both chemical and physical errors. The bulk of the results on the binary alloys in the systems iron-nickel, nickel-chromium, and chromium-iron may be taken from Dench's paper. The heats of formation of ironnickel alloys obtained by Steiner and Krisement ( 9 ) , who applied liquid-tin solution calorimetry a t 800" C., agree reasonably well with those of Dench and should also be taken into account. All this information has been used to construct smoothed isoenthalpy contours a t 500 cal. per gram-atom intervals, shown in Figure 1, obtained by graphical interpolation of the tabulated results. Figure 1 shows that the isoenthalpy curves show a relatively simple pattern. The maximum heat effect seems to occur in the ternary range, a t a composition of about 60 atomic 970 Cr and 30 atomic % Ni, where 2200 calories are absorbed in the formation of 1 gram-atom of alloy.
Table I . Integral Heats of Formation from the Component Metals of Ternary Iron-Nickel-Chromium Alloys at 1292" C., and Binary Iron-Nickel Alloys at 1050" C. Atomic Per Cent Cr Fe Ni
0 0 0 0 0 10 10 10 '10 10 10 10 10 10
20 20 20 20 20 20 20 20 30 30
25 25 90 90 90 10 20 30 40 50 60 65 70 80 10
20 30 40 40 50 60 70 10 20
75 75 10 10 10 80 70 60 50 40 30 25 20 10 70 60 50 40 40 30 20 10 60 50
AH/, Gal., G.-Atom
-1060 -1045 -195 -290 -140 -650 -490 -740 -510 -450 -80 -170 -130 +250 -270 -80 -310 -30 -120 +lo0 +550 +810 +340 +360
VOL. 12, No. 3, JULY 1967
Atomic Per Cent Cr Fe Ni
.
30 30 30 30 30 30 40 40 40 40 40 50 50 50 50 60 60 60 60 65 70 70 80
30 30 40 40 50 60 10 20 30 40 50 10 20 30 40 10 20 30 30 10 10 20 10
40 4030 30 20 10 50 40 30 20 10 40 30 20 10 30 20 10 10 25 20 10 10
Gal., G.-Atom
+450 +410 +770 +970 +860 +860 +850 +950 +1340 +1700 +1610 +1440 +1650 +1810 +1680 +2270 +1850 +1620 +1300 +1940 +1800 +1610 +1730
-
Fc
Ni
"i
Figure 1. lsoenthalpy contours (in calories per gram-atom) in the system iron-nickel-chromium for the range of temperature covered by the experimental work Discussion of the significance of the enthalpy changes would require a more intimate knowledge of the equilibrium diagram, which can be worked out only when a complete set of free energy data is also available. Table I1 records the measured, true, heat capacities of a chromium-nickel alloy with 35.0 atomic % Cr as well as those for the pure metals, the values for which agree reasonably well with the ones calculated additively from the heat capacities of chromium and nickel selected by ~ C p (alloy) Hultgren et al. ( 5 ) . The values of A C = C p (metals) show that there is a substantial deviation from Neumann-Kopp's rule indicating that the heat of formation of Ni-Cr solid solutions depends on temperature. The original experimental data are indicated in Figure 2 . Hultgren and Land (4)measured heat contents of nickelchromium alloys with 2 to 11 atomic % ' Cr between 400" and 1500°K. by means of a diphenyl ether calorimeter. At the higher temperatures, they generally also found positive deviations from Neumann-Kopp's rule. Though smaller than the deviations for the 35% alloy, they increase with chromium content and are thus consistent with the present findings. Taylor and Hinton (10) studied the atomic heat of an alloy, "CrNia," and found some evidence of superlattice formation in that the Cp us. T curve showed the typical h shape with the maximum a t 814°K. At higher temperatures, A c p again assumes positive values.
Table II. True Heat Capacities of a Ni-Cr Alloy with 35 Atomic Yo Cr and the Corresponding Composite of Chromium and Nickel
Temp., " C .
900 950 loo0
1 nm _ __.
1100 1150 1200
Alloy, C p , Cal.1 C., G.-Atom 8.79 8.91 8.97 9.06
9.14 9.19 9.28
0.35 Cr + 0.65 N i C p , Cal./ 'C., G.-Atom 8.10 8.23 8.36 8.48 8.61 8.74 8.87 ~
~~
LCP,
Ca1.i " C., G.-Atom 0.69 0.68 0.62 0.57 0.53 0.46 0.42
419
9.5 r
t -I 7
c)
Table I l l . Experimental Conditions and Heats of the Order-Disorder Transformation of FeNis
t
E
. -
9.0
Specimen
Pretreatment: Annealing Temp., C. and Time, Hr.
1 2 3 4
490", 935 & 455", 720 435", 270 & 400", 550 490", 22 & 490" to 350" 550 490", 1030 & 490" to 350", 550
Alloy
Elements
0
O
Cp Measmt. Temperatures, L,, Cal./ *C. G.-Atom 443-608 503-569 506-603 503-584
647 357 596 667
d
0
l , , , / l / , , , / , , , , , j
transition-namely, 1000-at 75 atomic % Ni. In both cases, the temperature range in which the magnetic transformation occurs was entered, so that some magnetic energy was contained in the calorific value. ACKNOWLEDGMENT
The work described above was carried out a t the National Physical Laboratory. The authors acknowledge assistance rendered by E. L. Glasson. LITERATURE CITED
Dench, W.A., Natl. Phys. Lab. Met. Diu. Rept. No. 32, January 1966. Dench, W.A., Trans. Faraday SOC. 59, 1279 (1963). Dench, W.A., Kubaschewski, O., J . Iron Steel Inst. (London) 201, 140 (1963). Hultgren R., Land, C., Trans. Met. SOC.A I M E 215, 165 (1959). Hultgren, R., On, R.L., Anderson, P.D., Kelley, K.K., "Selected Values of Thermodynamic Properties of Metals and Alloys," Wiley, New York, 1963. Iida, S., J . Phys. SOC. Japan 7, 373 (1952). Kaya, S., Nakayama, M., 2. Physik 112, 420 (1939). Leech, P., Sykes, C., Phil. Mag. 27, 742 (7) (1939). Steiner, W., Krisement, O., Arch. Eisenhuettenw. 32, 701 (1961). Taylor, A,, Hinton, K.G., J . Inst. Metals 81, 169 (1952).
RECEIVED for review January 3, 1967. Accepted March 17, 1967.
420
JOURNAL
OF CHEMICAL A N D ENGINEERING DATA
